Kimi K2.5: Now Top 1 on the OSWorld leaderboard. 🏆 With its Computer Use capabilities, you can now build powerful agents that navigate and operate computer interface just like a human.

🔥⚡ JUST IN: Anthropic drops Claude Haiku 5.5, its fastest model yet, costing around 75% less to run than Haiku 4.5 while scoring 72.4% on OSWorld vs 15.7% for the old version.

🚨HOLY FUCK! Claude Opus 5.5 Benchmarks is Terrifying it absolutely destroys Fable 5.1 and GPT-6 Astra and the previous Opus 5 on these benchmarks 66.4% on Terminal-Bench 57.8% on CursorBench 54.4% on Frontier Code 1846 on GDPval-AA 67.7% on Humanity’s Last Exam 81.8% on OSWorld 2.0 89.0% on Chartography Anthropic Cooked something serious🔥

Against Qwen3.7 Max, Qwen reports gains across long-horizon and visual agent benchmarks: PaperBench rises from 64.8 to 93.0, OSWorld-Verified from 73.3 to 86.1, and Vision2Web from 42.1 to 69.0.

My essay at TheSequence today covers the topic of environments in AI post-training AI environments are becoming a core ingredient of model capability: executable worlds that generate trajectories and feedback, not just exams. This essay traces their path from Atari, OpenAI Gym, and AlphaZero through MuJoCo to the RLVR-driven post-training behind Tulu 3 and DeepSeek-R1, then surveys the new generation of agent environments, including SWE-Gym, WebArena, OSWorld, TheAgentCompany, Tau2-bench, and AgentMercury. Check it out:

Claude Haiku 5.5 just dropped, and the benchmark jump is HUGE. @AnthropicAI's smallest model beats GPT-6 Luna on every benchmark where both have a reported score in the release chart. 🔥 Compared with Haiku 4.5: Computer use: 15.7% → 72.4% on OSWorld 2.1, versus Luna's 48.9%. Reasoning: 10.2% → 45.9% on Humanity's Last Exam without tools. With tools, it hits 57.4%. Agentic coding: 0% → 39.2% on Terminal-Bench 4.0, versus Luna's 16.4%. FrontierCode: 46.4%, beating Luna's 42.4% and approaching Sonnet 5.5's 52.1%. Visual reasoning: 6.4% → 46.4% on Chartography, versus Luna's 29.1%. Knowledge work: 1620 on GDPval-AA and 1578 on AA-Briefcase, ahead of Luna's 1437 and 1336. Sonnet 5.5 still leads across the chart, especially on Terminal-Bench. But the leap from the previous Haiku is enormous, particularly for computer use and reasoning.

💡 That task the top model could definitely handle, except the per-task price never worked out. Most of us have shelved one or two of those. What has kept frontier models out of production is usually not their capability but their price. Writing code as an agent, reading through long documents, driving a computer — all of it burns tokens, and knowing the best model would solve it doesn't help when the unit cost blocks the rollout. GPT-6.1 Sol, just announced by OpenAI, answers that wall by making the same capability cheaper rather than by pushing capability higher. Against the top-end GPT-6 Astra, it matches the score on DeepSWE v1.1 for coding at one-fifth the cost, and on OSWorld 2.0 for computer use it closes the gap to 2.1 points while costing one-seventh. Its hallucination rate improves from 4.5% to 4.1%, near Astra's 4.0%. API pricing is $2 per million input tokens and $10 output, with cached input at $0.10 — a 95% discount off standard input. The shift shows up not in tasks that only the top model could solve, but in the ones shelved purely over cost. On Terminal-Bench Science 0.1, what runs $23.80 per task on Astra costs $5.47 here. Workloads parked for failing the math come straight back onto the table, and for agents rereading the same context every turn, that cached-input price compounds the effect. Astra still has its place where the last few points genuinely matter. For now, routing by task difficulty and unit cost looks like the practical design. Title: Introducing GPT-6.1 Sol URL: #GPT6# #LLM#

Xiaomi MiMo-V2.6 is now open—a native multimodal agent family built for large-scale reinforcement learning. 🚀📜 MIT License. 🤖 🏆 MiMo-V2.6-Pro scores 46 on the Artificial Analysis Intelligence Index and reaches 71.9 on DeepSWE v1.1, 89.9 on Terminal-Bench 2.1, and 82.0 on OSWorld-Verified. 🧠 The 1.02T MoE activates 42B parameters and supports text, images, video, audio, and a 1M-token context. ⚙️ One mixed RL run trains coding, general, visual, and cybersecurity agents together. Pro and Flash completed 30 steps each in under six days, producing around 750K trajectories. 🌐 The models support computer use, 3D creation, embodied control, coding, design, video, and music workflows.

Introducing GPT-6 Sol and Luna, bringing the advances behind Astra to faster, more affordable models. ✨ 💻 Stronger coding and computer use 🎯 Improved factuality and alignment 💬 Clearer answers with less jargon > On AutomationBench, Sol at xhigh effort outperforms Claude Opus 5 at max effort at just 9% of its cost per task. > On OSWorld 2.0 offline, Luna at max effort exceeds GPT-5.6 Sol at medium effort at one-tenth the cost. > Sol makes about half as many mistakes as GPT-5.6 Sol on our internal factuality evaluation of conversations where users previously flagged errors. > Improved prompt caching helps agents reuse more context, with 90% discounts on cached input-token reads. Developers can change reasoning effort and tool availability without breaking cache. Rolling out today in ChatGPT Work and Codex for Plus, Pro, Business, Enterprise and Edu users, and available in the OpenAI API. Free and Go users can access Luna in the desktop app. These models are not yet available in Chat. API pricing per 1M tokens: Sol: $2 input / $10 output Luna: $0.10 input / $0.50 output

ANTHROPIC LAUNCHES CLAUDE OPUS 5.5, FIRST MODEL IN NEW 5.5 FAMILY Anthropic says Opus 5.5 performs at roughly the level of Claude Fable 5.1 on most tasks, while costing 40% less to run than Opus 5 and generating output 30%+ faster. Here’s everything you need to know: Opus 5.5 is Anthropic’s new flagship model, with major gains in agentic coding, computer use and knowledge work. Pricing: input_4M tokens • Output: $20/M • Cache reads: $0.20/M, 60% cheaper than Opus 5 Benchmark highlights: • Terminal-Bench 4.0: 66.4% vs. 52.3% for Opus 5 • CursorBench 4.0: 57.8% vs. 46.6% • GDPval-AA: 1846 vs. 1708 • OSWorld 2.0: 81.8% vs. 74.0% Anthropic says Opus 5.5 is especially strong on long-running coding work. In one test, it audited and fixed a 200K-line codebase in under 3 hours, versus more than 20 hours for Opus 5. Safety also improved. Anthropic says Opus 5.5 scored best among its recent models on its behavioral alignment audit and attempted to cross containment boundaries about 85% less often than Opus 5 or Mythos 5.1. Opus 5.5 is available now across Claude, AWS, Google Cloud and Microsoft Azure. Claude Sonnet 5.5 and Haiku 5.5 are coming in the next few weeks.